Details
Zusammenfassung: <jats:p>Some evidence suggests that Ca and vitamin D supplements affect cancer risk; however, it is uncertain whether the effects are due to Ca, vitamin D or the combination. We investigated the effect of Ca supplements without co-administered vitamin D on cancer risk. Medline, Embase and the Cochrane Central Register of Controlled Trials, reference lists of meta-analyses and two clinical trial registries were searched for randomised, placebo-controlled trials of Ca supplements ( ≥ 500 mg/d), with ≥ 100 participants and duration &gt;1 year. The lead authors of eligible trials supplied data on cancer outcomes. Trial-level data were analysed using random-effects meta-analyses and patient-level data using Cox proportional hazards models. A total of sixteen trials were eligible, six had no data available, ten provided trial-level data (<jats:italic>n</jats:italic> 10 496, mean duration 3·9 years), and of these, four provided patient-level data (<jats:italic>n</jats:italic> 7221, median duration 3·5 years). In the meta-analysis of trial-level data, allocation to Ca did not alter the risk of total cancer (relative risk 0·95, 95 % CI 0·76, 1·18, <jats:italic>P</jats:italic>= 0·63), colorectal cancer (relative risk 1·38, 95 % CI 0·89, 2·15, <jats:italic>P</jats:italic>= 0·15), breast cancer (relative risk 1·01, 95 % CI 0·64, 1·59, <jats:italic>P</jats:italic>= 0·97) or cancer-related mortality (relative risk 0·96, 95 % CI 0·74, 1·24, <jats:italic>P</jats:italic>= 0·75), but reduced the risk of prostate cancer (relative risk 0·54, 95 % CI 0·30, 0·96, <jats:italic>P</jats:italic>= 0·03), although there were few events. The meta-analysis of patient-level data showed similar results, with no effect of Ca on the risk of total cancer (hazard ratio 1·07, 95 % CI 0·89, 1·28, <jats:italic>P</jats:italic>= 0·50). Ca supplements without co-administered vitamin D did not alter total cancer risk over 4 years, although the meta-analysis lacked power to detect very small effects, or those with a longer latency.</jats:p>
Umfang: 1384-1393
ISSN: 0007-1145
1475-2662
DOI: 10.1017/s0007114513001050